Turning force tracking improvement using a flatness-based predictive approach

Generalized predictive control has been introduced by Clarke et al. and consisted in finding a control by minimization of a linear quadratic cost function with receding horizon. A prediction method for control calculation based on the partial state concept was presented by M´Saad et al. and allowed the direct specification of desired performances for tracking and for the regulation dynamics. The aim of this paper is to present these two predictive techniques and to propose some new ideas permitting the obtention of high performances level in terms of a tracking by the use of the flatness property. The proposed strategy is applied to the tracking of a turning force. Simulations results are given and show the improvement in the tracking obtained by the use of flatness.